Molecular Dynamics Simulations of Ionic Transport in Molten Salt Systems

Molten salt systems, comprising molten chlorides, fluorides and multicomponent mixtures, are pivotal in advanced energy applications, including nuclear reactors, concentrated solar power and thermal energy storage. Molecular dynamics simulations offer atomistic insight into ionic structure, speciation and transport by explicitly modelling interionic forces over femtosecond time-scales. Classical approaches employ rigid-ion or polarizable-ion potentials to capture short- and long-range interactions, while ab initio molecular dynamics (AIMD) derives forces on the fly from electronic structure methods, trading speed for accuracy. Recent advances in machine-learning interatomic potentials bridge this gap, delivering near-quantum accuracy with orders-of-magnitude speed-ups, enabling large-scale, long-time simulations of complex molten salt chemistries. Such studies elucidate coordination environments, network formation and self-diffusion coefficients, directly informing thermophysical property predictions, ionic conductivity estimates and corrosion-control strategies. Together, these methods provide a comprehensive toolkit for understanding and optimising molten salt behaviour under reactor-relevant temperatures, compositions and redox conditions.

Research from Nature Portfolio

Contemporary reviews chart the evolution of computational methodologies for molten salts, dividing progress into three eras: early rigid-ion models, the advent of polarizable potentials and the current rise of machine-learning frameworks. The latest work synthesises seven decades of developments, critically assessing the strengths and limitations of each approach in predicting densities, heat capacities and ionic transport properties. It highlights how emerging neural-network potentials and hybrid quantum-mechanics/machine-learning schemes are set to overcome the trade-off between accuracy and efficiency. This forward-looking perspective underscores the potential for rapid, high-throughput screening of novel salt chemistries and the integration of these methods into reactor design workflows.

Molecular Dynamics Simulations of Ionic Transport in Molten Salt Systems publication trend

The graph below shows the total number of articles in molecular dynamics simulations of ionic transport in molten salt systems across all publications each year (not limited to Nature Index journals).

Technical terms

Molecular dynamics simulation: A computational method that calculates the time evolution of a system of particles by integrating Newton’s equations of motion.

Ab initio molecular dynamics (AIMD): A form of molecular dynamics in which interatomic forces are computed directly from quantum mechanical calculations at each time step.

Neural network interatomic potential (NNIP): A machine-learning model that maps atomic environments to potential energies and forces, trained on quantum mechanical reference data.

Polarizable ion model (PIM): An interatomic potential that allows ionic charges or dipoles to adjust dynamically in response to the local electric field.

Self-diffusion coefficient: A measure of the rate at which individual ions migrate through the molten salt via thermal motion, typically expressed in m²·s^–1.

References

1. Development of robust neural-network interatomic potential for molten salt. Cell Reports Physical Science (2021).
2. Modeling LiF and FLiBe Molten Salts with Robust Neural Network Interatomic Potential. ACS Applied Materials & Interfaces (2021).
3. Computational methods to simulate molten salt thermophysical properties. Communications Chemistry (2022).
4. The impact of hydrogen valence on its bonding and transport in molten fluoride salts. Journal of Materials Chemistry A (2021).